Merge pull request #178 from static-frame/177/np2

NumPy 2.0 support

Author: flexatone

performance/__main__.py

@@ -507,9 +507,9 @@ def __init__(self):
         self.values = [
                 np.longlong(-1), np.int_(-1), np.intc(-1), np.short(-1), np.byte(-1),
                 np.ubyte(1), np.ushort(1), np.uintc(1), np.uint(1), np.ulonglong(1),
-                np.half(1.0), np.single(1.0), np.float_(1.0), np.longfloat(1.0),
-                np.csingle(1.0j), np.complex_(1.0j), np.clongfloat(1.0j),
-                np.bool_(0), np.str_('1'), np.unicode_('1'), np.void(1),
+                np.half(1.0), np.single(1.0), np.float64(1.0), np.longdouble(1.0),
+                np.csingle(1.0j), np.complex_(1.0j), np.clongdouble(1.0j),
+                np.bool_(0), np.str_('1'), np.str_('1'), np.void(1),
                 np.object(), np.datetime64('NaT'), np.timedelta64('NaT'), np.nan,
                 12, 12.0, True, None, float('NaN'), object(), (1, 2, 3),
                 NT(1, 2, 3), datetime.date(2020, 12, 31), datetime.timedelta(14),

requirements-build-3_11.txt

@@ -1,2 +1,2 @@
-numpy==1.23.5
+numpy==2.0.0
 

requirements-build-3_12.txt

@@ -1,2 +1,2 @@
-numpy==1.26.2
+numpy==2.0.0
 setuptools==69.*

setup.py

@@ -41,8 +41,8 @@ def get_ext_dir(*components: tp.Iterable[str]) -> tp.Sequence[str]:
             'src/methods.c',
             'src/tri_map.c',
         ],
-        include_dirs=get_ext_dir('numpy', 'core', 'include'),
-        library_dirs=get_ext_dir('numpy', 'core', 'lib'),
+        include_dirs=get_ext_dir('numpy', '_core', 'include'),
+        library_dirs=get_ext_dir('numpy', '_core', 'lib'),
         define_macros=[("AK_VERSION", AK_VERSION)],
         libraries=['npymath'], # not including mlib at this time
         )

src/delimited_to_arrays.c

@@ -1069,7 +1069,7 @@ AK_CPL_Free(AK_CodePointLine* cpl)
     PyMem_Free(cpl->buffer);
     PyMem_Free(cpl->offsets);
     if (cpl->type_parser) {
-        PyMem_Free(cpl->type_parser);
+        AK_TP_Free(cpl->type_parser);
     }
     PyMem_Free(cpl);
 }
@@ -1364,7 +1364,7 @@ AK_CPL_to_array_float(AK_CodePointLine* cpl, PyArray_Descr* dtype, char tsep, ch
     // initialize error code to 0; only update on error.
     int error = 0;
     bool matched_elsize = true;
-    int elsize = dtype->elsize;
+    int elsize = PyDataType_ELSIZE(dtype);
 
     NPY_BEGIN_THREADS_DEF;
     NPY_BEGIN_THREADS;
@@ -1442,7 +1442,7 @@ AK_CPL_to_array_int(AK_CodePointLine* cpl, PyArray_Descr* dtype, char tsep)
     // initialize error code to 0; only update on error.
     int error = 0;
     bool matched_elsize = true;
-    int elsize = dtype->elsize;
+    int elsize = PyDataType_ELSIZE(dtype);
 
     NPY_BEGIN_THREADS_DEF;
     NPY_BEGIN_THREADS;
@@ -1515,7 +1515,7 @@ AK_CPL_to_array_uint(AK_CodePointLine* cpl, PyArray_Descr* dtype, char tsep)
     // initialize error code to 0; only update on error.
     int error = 0;
     bool matched_elsize = true;
-    int elsize = dtype->elsize;
+    int elsize = PyDataType_ELSIZE(dtype);
 
     NPY_BEGIN_THREADS_DEF;
     NPY_BEGIN_THREADS;
@@ -1583,15 +1583,15 @@ AK_CPL_to_array_unicode(AK_CodePointLine* cpl, PyArray_Descr* dtype)
     bool capped_points;
 
     // mutate the passed dtype as it is new and will be stolen in array construction
-    if (dtype->elsize == 0) {
+    if (PyDataType_ELSIZE(dtype) == 0) {
         field_points = cpl->offset_max;
-        dtype->elsize = (int)(field_points * UCS4_SIZE);
+        // dtype->elsize = (int)(field_points * UCS4_SIZE);
+        PyDataType_SET_ELSIZE(dtype, (npy_intp)(field_points * UCS4_SIZE));
         capped_points = false;
     }
     else {
         // assume that elsize is already given in units of 4
-        // assert(dtype->elsize % UCS4_SIZE == 0);
-        field_points = dtype->elsize / UCS4_SIZE;
+        field_points = PyDataType_ELSIZE(dtype) / UCS4_SIZE;
         capped_points = true;
     }
 
@@ -1649,13 +1649,14 @@ AK_CPL_to_array_bytes(AK_CodePointLine* cpl, PyArray_Descr* dtype)
     Py_ssize_t field_points;
     bool capped_points;
 
-    if (dtype->elsize == 0) {
+    if (PyDataType_ELSIZE(dtype) == 0) {
         field_points = cpl->offset_max;
-        dtype->elsize = (int)field_points;
+        // dtype->elsize = (int)field_points;
+        PyDataType_SET_ELSIZE(dtype, (npy_intp)field_points);
         capped_points = false;
     }
     else {
-        field_points = dtype->elsize;
+        field_points = PyDataType_ELSIZE(dtype);
         capped_points = true;
     }
 

src/methods.c

@@ -560,16 +560,16 @@ isna_element(PyObject *m, PyObject *args, PyObject *kwargs)
     }
     if (PyArray_IsScalar(element, Complex64)) {
         npy_cfloat val = PyArrayScalar_VAL(element, Complex64);
-        return PyBool_FromLong(isnan(val.real) || isnan(val.imag));
+        return PyBool_FromLong(isnan(npy_crealf(val)) || isnan(npy_cimagf(val)));
     }
     if (PyArray_IsScalar(element, Complex128)) {
         npy_cdouble val = PyArrayScalar_VAL(element, Complex128);
-        return PyBool_FromLong(isnan(val.real) || isnan(val.imag));
+        return PyBool_FromLong(isnan(npy_creal(val)) || isnan(npy_cimag(val)));
     }
     # ifdef PyComplex256ArrType_Type
     if (PyArray_IsScalar(element, Complex256)) {
         npy_clongdouble val = PyArrayScalar_VAL(element, Complex256);
-        return PyBool_FromLong(isnan(val.real) || isnan(val.imag));
+        return PyBool_FromLong(isnan(npy_creall(val)) || isnan(npy_cimagl(val)));
     }
     # endif
 

src/tri_map.c

@@ -14,7 +14,9 @@
 static inline NPY_DATETIMEUNIT
 AK_dt_unit_from_array(PyArrayObject* a) {
     // This is based on get_datetime_metadata_from_dtype in the NumPy source, but that function is private. This does not check that the dtype is of the appropriate type.
-    PyArray_DatetimeMetaData* dma = &(((PyArray_DatetimeDTypeMetaData *)PyArray_DESCR(a)->c_metadata)->meta);
+    PyArray_Descr* dt = PyArray_DESCR(a); // borrowed ref
+    PyArray_DatetimeMetaData* dma = &(((PyArray_DatetimeDTypeMetaData *)PyDataType_C_METADATA(dt))->meta);
+    // PyArray_DatetimeMetaData* dma = &(((PyArray_DatetimeDTypeMetaData *)PyArray_DESCR(a)->c_metadata)->meta);
     return dma->base;
 }
 
@@ -856,9 +858,9 @@ AK_TM_fill_object(TriMapObject* tm,
 #define AK_TM_TRANSFER_FLEXIBLE(c_type) do {                               \
     Py_ssize_t one_count = from_src ? tm->src_one_count : tm->dst_one_count;\
     TriMapOne* one_pairs = from_src ? tm->src_one : tm->dst_one;           \
-    npy_intp t_element_size = PyArray_DESCR(array_to)->elsize;             \
+    npy_intp t_element_size = PyArray_ITEMSIZE(array_to);                  \
     npy_intp t_element_cp = t_element_size / sizeof(c_type);               \
-    npy_intp f_element_size = PyArray_DESCR(array_from)->elsize;           \
+    npy_intp f_element_size = PyArray_ITEMSIZE(array_from);                \
     c_type* array_to_data = (c_type*)PyArray_DATA(array_to);               \
     c_type* f;                                                             \
     c_type* t;                                                             \
@@ -906,7 +908,7 @@ AK_TM_fill_unicode(TriMapObject* tm,
 
     Py_UCS4* array_to_data = (Py_UCS4*)PyArray_DATA(array_to);
     // code points per element
-    npy_intp cp = PyArray_DESCR(array_to)->elsize / UCS4_SIZE;
+    npy_intp cp = PyArray_ITEMSIZE(array_to) / UCS4_SIZE;
 
     bool decref_fill_value = false;
     if (PyBytes_Check(fill_value)) {
@@ -948,7 +950,7 @@ AK_TM_fill_string(TriMapObject* tm,
             ? tm->final_src_fill : tm->final_dst_fill);
 
     char* array_to_data = (char*)PyArray_DATA(array_to);
-    npy_intp cp = PyArray_DESCR(array_to)->elsize;
+    npy_intp cp = PyArray_ITEMSIZE(array_to);
     if (!PyBytes_Check(fill_value)) {
         return -1;
     }

test/test_util.py

@@ -527,11 +527,11 @@ def test_dtype_from_element_core_dtypes(self) -> None:
                 np.ulonglong,
                 np.half,
                 np.single,
-                np.float_,
-                np.longfloat,
+                np.float64,
+                np.longdouble,
                 np.csingle,
-                np.complex_,
-                np.clongfloat,
+                np.complex128,
+                np.clongdouble,
                 np.bool_,
         ]
         for dtype in dtypes:
@@ -540,12 +540,12 @@ def test_dtype_from_element_core_dtypes(self) -> None:
     def test_dtype_from_element_str_and_misc_dtypes(self) -> None:
         dtype_obj_pairs = [
                 (np.dtype('<U1'), np.str_('1')),
-                (np.dtype('<U1'), np.unicode_('1')),
+                (np.dtype('<U1'), np.str_('1')),
                 (np.dtype('V1'), np.void(1)),
                 (np.dtype('O'), object),
                 (np.dtype('<M8'), np.datetime64('NaT')),
                 (np.dtype('<m8'), np.timedelta64('NaT')),
-                (np.float_, np.nan),
+                (np.float64, np.nan),
         ]
         for dtype, obj in dtype_obj_pairs:
             self.assertEqual(dtype, dtype_from_element(obj))